Tufts GIS Center

1

Downloading Census Data from American Factfinder for use in ArcGIS

Written by Barbara Parmenter, revised September 24 2013

OBTAINING DATA FROM AMERICAN FACTFINDER (AFF) .......................................................................... 1

PREPARING AMERICAN FACTFINDER DATA FOR USE IN ARCMAP ............................................................ 5

OBTAINING GIS FILES FOR CENSUS GEOGRAPHY ...................................................................................... 9

JOINING THE AFF TABLE TO YOUR CENSUS TRACT POLYGONS IN ARCMAP ............................................ 11

SETTING A PROJECTED COORDINATE SYSTEM FOR YOUR MAP .............................................................. 16

In this tutorial, we will be obtaining information about housing tenure at the Census Tract level from the 2010 Census for a single county using American Factfinder. You can then use a similar process to download any other Census 2010, American Community Survey, or Census 2000 data for other geography levels and/or for whole states or multiple counties. You have many, many options in American Factfinder – this shows one possible path.

Obtaining Data from American FactFinder (AFF) 1. Data management is critical when dealing with the multiple tables of the Census. Before

beginning this tutorial: a. Create a Census 2010 folder in your personal workspace b. Create two subfolders: AFF Data and Census Geography

2. Go to the US Census web site – http://census.gov and under the Data tab select

American FactFinder

Tufts GIS Center

2

3. Click on Advanced Search (and if you see it, Show Me All) 4. Click on Topics in the left column:

5. Click on Dataset and scroll down to click on 2010 SF1 100% Data – this will send this

criteria to your Selection box in the upper left of the site:

Tufts GIS Center

3

6. Scroll up in the Topics list and click on Housing – then under Occupancy Characteristic, click on Owner/Renter (Tenure in Occupied Units)

7. Close the Topics box (see above graphic) 8. Click on Geographies on the left column – this brings up the Select Geographies overlay 9. Fill out the box so that you are selecting all Census Tracts for a specific state and a

county in that state – below we are selecting all census tracts in Suffolk County, Massachusetts

Tufts GIS Center

4

10. Be sure to click on ADD TO YOUR SELECTIONS 11. Close the Select Geographies overlay

12. Be sure that the Your Selections box in the upper left corner contains what you want – the data set, the general topic, and the census geography level for the specific location you want (all tracts, not just one tract). If it does not say this, clear your selections and start over from Step 3 above.

Tufts GIS Center

5

13. Checkmark a table of interest – to see what variables a table contains, click on the Information icon for that table. For this exercise, we highly recommend a table with just a few variables, for example, H11. TOTAL POPULATION IN OCCUPIED HOUSING UNITS BY TENURE – this will show you how many people live in rented units, units owned through a mortgage, and units owned free and clear or without payment. This is an easy table to process in Excel.

14. Click on Download ( ) and follow the instructions (your file will be built, and then you will click on another Download option)

15. The file is in compressed format – extract it to your Census 2010/AFF Data folder

Preparing American Factfinder Data for Use in ArcMap Double-click on both downloaded .csv files to open them in Excel:

Note: If you are opening the file from within Excel, you will need to set the option to look for all file types:

Tufts GIS Center

6

Your DEC_10…with_ann.csv file should look something like this – this file contains the data:

Now open the file - DEC_10… _metadata.csv file. This file explains the column header codes in the data file - it should look something like what you see below. This is a very important file!!! Typically the first data column (D001 here) is the Universe of things counted in this table. This table is counting people in occupied housing units. In the case of this table if you wanted to show the % of the population that is in rented housing units, you would divide D004 by D001 and multiply by 100:

A few important steps left. First, ArcGIS does not like periods in the column names: 1. Delete all periods (.) in all the column names

Second, the Geo ID in the Census Tracts polygon attribute table to which you will be joining this data table is in a text format. GEOID2 in this CSV file must also be text for the join to work

properly.

2. Click on the tab (B) above GEO.ID2 to highlight the entire column

Tufts GIS Center

7

3. Click on the Excel Tab for Data, then click on Text to Columns:

4. Click Next to leave the first setting at Delimited 5. Click Next to leave the second setting at Tab

6. On Step 3, change the column data format to TEXT:

7. To make things easier later, rename the worksheet to something comprehensible, e.g.,

Housing_Tenure - the worksheet name will be the identifier (what is displayed) in

ArcCatalog

8. Very important step – save your modified CSV file as an Excel file (.xlsx) – give it a

comprehensible name, e.g., 2010 Census H11_population by housing tenure.xlsx Optional Tip – you can change your column names to match what the metadata .csv file is

telling you, or you can keep the metadata file handy so that you know what the codes stand

Tufts GIS Center

8

for when you are ready to use this data in ArcGIS. It’s probably easier to refer to the codes

later. Extra step for Alaska, Alabama, Arkansas, Arizona, California, Colorado, and Connecticut

(if you’re not getting data for one of these states, ignore this section) Some states have FIPS codes that start with a zero, and because Excel removes that leading zero, the table won't join properly unless we put it back on. If you're working in Alaska, Alabama, Arkansas, Arizona, California, Colorado, or Connecticut, you'll have to add that zero back on manually.

To add the zero back on, at the beginning of the table, under Column A, at Row2 type in: =concatenate(“0”,B2) That's a zero inside the quote

Hit Enter.

If the result of that formula looks right (i.e. it has a leading zero), copy that cell’s formula to the rest of the column.

You're almost done! To keep this compatible with the rest of the directions, copy all of the cells in this new column, and right click on the GEOid2 column. Click the options below “Paste Special”, choose Values, and your leading zeroes should be all set. Ensure your

column is still named GeoID2.

Delete the column you added but be sure you still have the fixed GeoID2 column. 9. Note: your table may have columns, like D001, where the data has text values in it. This

will cause this information to not be mappable in ArcMap. Follow the directions below if

this is the case:

Tufts GIS Center

9

a. Delete any information in parentheses

b. So your table should now look l like this:

Save your Excel file again!

Final STEP: Save your file EXIT out of Excel – you CANNOT have Excel open still when you work with this data in ArcGIS!

Obtaining GIS files for Census Geography Now you need to get your Census tract polygons…

1. Go to the Census web site (http://census.gov) and click on the Geography tab and select the option TIGER

Tufts GIS Center

10

2. Click on Tiger/Line Shapefiles and Files in the TABLE as shown below:

3. Click on the link for the 2013 Tiger/Line Shapefile Main Page

4. Click on Download Shapefiles then on Web Interface

5. Under Select a Layer Type use the pulldown menu to choose Census Tracts then press Submit

6. Select your State of interest and download the data set – it is compressed in a .zip file 7. Unzip the downloaded file to your Census Geography folder

Tufts GIS Center

11

Joining the AFF table to your Census Tract polygons in ArcMap Now you’re ready for mapping!

1. Start a session of ArcMap with a blank map 2. Add your Census Tracts geography data set to the map (e.g., tl_2012_25_tract.shp) 3. Rename your Census Tract geography layer to Census Tracts 4. Open the Census Tracts polygon attribute table and take a look at it 5. The GeoID column is what we will be using for joining our AFF data 6. Right-click on GEOID field name and choose Properties

You’ll see it is a STRING type attribute field – that’s important to know. Close the table.

7. Add your American Factfinder table to the map - you need to drill down to the worksheet level:

8. Open the American Factfinder Table by right-clicking on it and choosing Open

Tufts GIS Center

12

9. GEOid2 will be your join field – check its properties to ensure that it is also a STRING

type and that all the census data appears correctly 10. Close the table 11. Right click on your Census Tracts and choose Join and Relates – Join…

Tufts GIS Center

13

12. Fill in the dialog box as follows – you are joining attributes from a table, using GEOID in

your Census Tracts layer and GEOid2 in your AFF table – click OK when done:

13. Open the Census Tracts attribute table to ensure that the join was made correctly. If it was, you should see your AFF data when you scroll to the right in the table.

14. Close the table

Tufts GIS Center

14

One tip – if your area of interest is near water or has water features in it (like Boston), your tract data set may look like this – it includes a lot of water:

To get rid of the water tracts when you make a map, in the Symbology properties, click on Quantities and choose your variable, but then also click on Classify.

In the Classify dialog box, click on Exclusion. You can exclude all census tracts where the land area field = 0 (no land), as follows:

Tufts GIS Center

15

Now you can make a map of your data following the usual methods. If you are unfamiliar with mapping numeric values, see the ArcGIS 10.1 online help – About Symbolizing Layers to Represent Quantity

The next page shows an example of a map showing the % of people in rental housing units for each tract in Suffolk County (population renting normalized by total population in housing units). The map is using the “Light Gray Canvas” option from ESRI’s basemap choices (Click on File – Add Data – Add Basemap to get this option):

Here is the symbology properties for the map above:

Tufts GIS Center

16

Setting a Projected Coordinate System for your Map It is good cartographic practice to put your map into a projected coordinate system. The TIGER data is in a geographic coordinate system and can appear stretched in an odd way on your map. You can fix this problem by setting a projected coordinate system appropriate for your region.

You will need to know the best coordinate system to use for your area. In the case of Massachusetts, we will use the Massachusetts State Plane (NAD83) – meters coordinate system. If you don’t know what coordinate system to use, you can leave your map as is or ask a lab assistant.

1. Click on View – Data Frame Properties 2. Click on the Coordinate System tab 3. In the dialog box, scroll to find the Projected Coordinate Systems folder and the State

Plane folder within – click on that folder and select NAD1983 (meters) from the list:

Tufts GIS Center

17

4. Find NAD_1983_StatePlane_Massachuestts Mainland (not Islands!) and click on it

5. Click OK 6. Click Yes when warned that the coordinate system is different from the data in your

maps.

You’re done!